Jump to content

Retrospect Problem with Scheduled Backups


cpguru21

Recommended Posts

Hi all!

 

I am having an issue. Came in this morning, connected to my server only to find Retrospect stuck (i guess?) in process of a backup................

 

This is the last entry in my log:

 

" Completed: 137 files, 36.7 MB, with 55% compression

Performance: 26.0 MB/minute (14.6 copy, 122.2 compare)

Duration: 00:19:49 (00:17:00 idle/loading/preparing)

 

1/11/2010 7:27:19 PM: Execution completed successfully

Total performance: 580.1 MB/minute with 74% compression

Total duration: 00:27:09 (00:20:57 idle/loading/preparing)

Exit at 1/12/2010 10:23 AM

 

+ Retrospect version 8.1.626.1

Launched at 1/12/2010 10:24 AM"

 

Notice above where on the 11th, at 7:27:19 pm execution completed.....but then exited the next morning? Then Launched a backup again at 10:24AM? What the .........

 

Server osx 10.6.2

Retrospect Ver 8 (626)

 

Backups:

nightly backups - backups from the script menu as I understand are incremental.

 

Schedule is every day.

 

Time is 7:00pm every evening.

 

I do have thorough verification turned on and data compression (in software)

 

I just got this backup working, has been successfull with this configuration for approx 2 weeks now.

 

Really getting tired of 2 well 3 things....

 

1.Issues with retrospect (i know general but look at 2)

2.USELESS logs that show nothing if a backup is unsuccessfull. should at least LOG backup such and such has been running for such and such....

3.calling only to find out that this is some unresolvable issue that I have to reboot/restart engine and or server to resolve. all i can say is WOW.

and a 4. at the lack of manual so a newb like me has to get this frustrated with a product my company depends on.....

Link to comment
Share on other sites

+ Retrospect version 8.1.626.1

Launched at 1/12/2010 10:24 AM"

 

 

 

This (to me) means that this is when the *engine* started. If you didn't manually stop/start the engine -- then the engine crashed and automatically relaunched.

 

However, in my experience, the engine doesn't automatically relaunch under 10.6, so I would think the engine was manually stopped/started (did the server reboot?)

 

You should check your /Library/Logs/CrashReporter and/or /Library/Logs/DiagnosticReports folders to see if there's a crash log in them and submit this if you can reproduce the crash.

Link to comment
Share on other sites

It is possible that the server was re-booted at this time. It was not manually restarted.

 

Checked crash log and nothing. I will check the other log you mentioned in your post.

 

As far as reproducing....im not sure how. System has been running fine for my needs for approx 2 weeks now. May have to wait another 2 weeks for this to happen again.

 

Thanks!

Link to comment
Share on other sites

It is possible that the server was re-booted at this time. It was not manually restarted.

Well, if the Retrospect engine was doing something important (grooming, writing to a backup set) when the re-boot occurred, you could have caused problems there.

 

You also don't say whether it was a forced reboot or a graceful restart. A forced reboot / shutdown can give Retrospect fits because the sanity-checking logic doesn't seem to be very mature, causing crashes if the data isn't perfect.

 

Russ

Link to comment
Share on other sites

To me -- the fact that there's an *exited* line means (to me) that it didn't crash, but that the engine was stopped intentionally.

 

When I have engine crashes, there are no "exit" lines in the logs -- just a follow up "launched" line for when it was restarted.

 

 

Apart from that, from the Original Post, I'm not sure why you think things were "stuck"?

Link to comment
Share on other sites

Russ,

 

thanks for reply...always look forward to your input on these forums.

 

Shutdown was a planned shutdown. I shutdown the server for maintenance.

 

Maser,

again thanks for replying.

 

Maybe we are focusing on the exit in the log. MY inexperiance with retro may lead me astray from the issue.

 

The issue is when I connected to server and looked at the GUI, my backup script was stuck on "preparing to execute.."

 

The launch time of the backup was incorrect. Last backup registered from 1/11/10....should be one for every day.

 

I am turning engine logging on and hope the error can get re-created and get some real log info...

 

thanks!

Link to comment
Share on other sites

I've seen the "preparing to execute..." stick here on an activity (though not in a while, actually...)

 

 

If you get that, the first thing to do is to quit the GUI wait about 5 minutes and relaunch it to see if it's moved on from that (I've seen that happen here where the console stops refreshing).

 

If it still hasn't, then I've shut down the engine from the System Preference. NOTE: It usually takes about 10 minutes for the engine to actually *stop* in these instances. Wait for it...

 

 

 

I've submitted a number of Shark reports on these "preparing to execute..." sticks, so I think they are looking into why this is happening and I'm hopeful that the upcoming release will be addressing that.

 

 

Edited by Guest
Link to comment
Share on other sites

I understand Maser.

 

I am trying to be real patient with Retro, as I am a new install and so far seemed pretty straight forward to set up even for someone new to this software.

 

Well I have stopped and restarted the service. I am now going to go back in and re-set so my backups start at the correct time again. (by reset i mean check the activity and make sure backup is scheduled for tonight)

 

Its just that this problem started on the 11th or the 12th. No log anywhere shows that the program had an issue since then. We should not have to turn debbugging on to see an error if the script stoped unexpectidly or had to re-start. All I see is from the retro log:

+ Retrospect version 8.1.626.1

Launched at 1/12/2010 10:24 AM"

then nothing anywhere in the system except that activity shows a backup started "Today at 9:25"

 

Need better error reporting.

 

Cant wait for the updates coming out this month.

Edited by Guest
add more info for re-set
Link to comment
Share on other sites

I guess you could set up the e-mail notification on it? That way if you don't get the nightly e-mail that the backup finished, then you know there would be a problem?

 

But, I agree, the program needs to be able to run 100% of the time when there aren't obvious issues that need to be manually addressed by an admin (such as a full media set.)

Link to comment
Share on other sites

I have e-mail notification set up, but I switched e-mail and didnt set my "log" account Im sending mail to up, so I missed a couple of days.

 

Anyway stopping, then starting the engine did the trick. Im now off and running. I have debugging turned on so when this happens again, when I call support, maybe they can look into why this happens every so often.

 

Thanks again for your help!

Link to comment
Share on other sites

  • 2 weeks later...
  • 3 weeks later...

so its monday morning. I love mondays. No really.

 

1st part of my day is reviewing backup logs for the weekend. My backups are not sooo critical that I need to monitor over the weekend.

 

I noticed that my server Aphrodite (or specifically the retro running on it) did not send me a log since Friday. Hmmm.

 

Connected to server. Checked activity monitor and retro engine is running. Launched retro and wammmo!!! my backup starts running with a time stamp of when I opened the program. 8:13am. This morning. Hmmmm. Backup is scheduled for 19:00:00. Getting so frustrated with these issues. WTF are the updates?

 

I started reboot servers once a week to try to resolve these issues, and seemed to work for a month. Well i turned on ?debugging? (not sure what its called) for the logs so now I can call Retro support and have them tell me they arnt really sure whats going, but hold your breath for the update!

 

well im cancelling backup. I know from experiance, I can just cancel and then tonight, backup will run. as a precaution, I will restart the engine to make sure.

 

 

Here is a snipit of my log....I will report back after i call retro support.......

 

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4 bytes

MemAlloc: caller requested 4100 bytes

MemAlloc: caller requested 12288 bytes

MemAlloc: caller requested 20480 bytes

MemAlloc: caller requested 28672 bytes

MemAlloc: caller requested 36864 bytes

MemAlloc: caller requested 36864 bytes

MemAlloc: caller requested 16 bytes

MemAlloc: caller requested 9836 bytes

MemAlloc: caller requested 16 bytes

MemAlloc: caller requested 9836 bytes

MemAlloc: caller requested 16 bytes

MemAlloc: caller requested 9836 bytes

MemAlloc: caller requested 16 bytes

MemAlloc: caller requested 9836 bytes

TMacVol::GetInfo: vol /, blocksize [4096|4096], total [120988852224|120988852224], vfree [60756643840|60756643840], used [60232208384|68822142976],

TMacVol::GetInfo: vol /Volumes/Cota_Server_Raid_Aphrodite, blocksize [4096|4096], total [999860895744|999860895744], vfree [819452440576|819452440576], used [180408455168|137458782208],

MemAlloc: caller requested 32 bytes

MemAlloc: caller requested 1056 bytes

MemAlloc: caller requested 8 bytes

MemAlloc: caller requested 177944 bytes

Link to comment
Share on other sites

well it seems I have left out some very important information in my posts:

 

I am backing up to a NAS device. This device is on a separate subnet, connected to a separate NIC on my server. Ethernet cable is run approx 100' to an office in position of if we cant get to server, we can smash and grab the NAS devices for restoration. Logic for the second NIC was so we could minimize traffic on the existing network. All backup traffic travels across a completely separate network.

 

According to the technician I talked to today, the issue may be the second NIC and retro depending on apple networking. I was told that this is a known issue, although im not sure issue is the right word? The technician gave me the impression that this is "par for the course" in my configuration. The fact that 1 server was not having this problem was great, but he was not surprised I was experiencing an issue on the other server.

 

Debugging turned on, was told that there was so much logging information that my logs were useless. We turned logging back to 3 from 6.

 

I am now re-creating the script (using same media sets etc..) _per the technician_ to make sure it is not the script, as it seemed like the engine was launching the script when I opened the console in the described issue here, and not on the schedule specified.

 

So I will re-create and see if the problem appears again. May have to wait 2 weeks or a month, but I will let you know.

 

Can anyone add "2 cents" on my configuration? Is this an abnormal way to perform the backup? I just wanted to minimize traffic on our main network.....

 

Anybody else backing up to NAS devices at all?

 

My NAS is Thecus N7700. This device so far is working great. I have 2, one for each server.

 

Thanks for any and all constructive criticism, suggestions etc...

 

 

Edited by Guest
_per the technician_ added
Link to comment
Share on other sites

that was according to the retro tech.

 

NAS is has a shared folder called backup. Shared through AFP.

 

My previous post in this thread was an off the cuff reaction to the "tech's" comment. I've just gone back over the entire thread and am at a loss to see where anything about networking is implicated.

 

If you are able to Add the NAS as a Source in Retrospect (either by IP address or fully qualified domain name) it shouldn't matter at all through what hardware or router tables the packets are traveling. All that stuff is at a lower level then Retrospect. And Apple's networking stack in OS X is at least acceptably robust, AFAIK.

 

The NAS configuration question I was asking was about protocol, which you document as AFP (although there are different versions of this). But again, what here makes networking a suspicion? Does Retrospect no longer see the NAS as a Source? Do Members stored on this Source remain accessible?

 

What led the tech to talk about networking at all?

Link to comment
Share on other sites

Dave,

 

I too am a litt.e thrown off when the tech mentioned networking as a possible issue.

 

What led the tech to discuss this was me explaining my configuration. He immediately mentioned that he had doubts about the stability of my setup, and mentioned he was not surprised I had problems....

 

sigh........

 

I dont know enough about the product, and am too trusting to the technician of a given product, whether its retro or apple or whatever.

 

I am not too concerned about the issue as it only happens every 2 weeks - 1 a month.

 

I have 1 server that has been backing up this way without a hitch. Again 2 nics. 1 connected to giga switch, and 1 conneced directly to NAS (i think i left that last part out as well)

 

thanks for taking the time to read and reply Dave.

 

*edit*

 

Dave Wrote:

"Does Retrospect no longer see the NAS as a Source? Do Members stored on this Source remain accessible?"

 

Im not sure I can answer this right now. I have not tried to access the source when retro has had the problem. I do know that stopping, then starting the engine, I can access everything.

 

 

Edited by Guest
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
×
×
  • Create New...